Researchers at San Diego State University have developed an artificial intelligence (AI) system to detect and expose cryptocurrency giveaway scams on Twitter.
Dubbed GiveawayScamHunter, the automated system has identified a staggering 95,111 scam lists created by 87,617 accounts on the social media platform between June 2022 and June 2023.
By using GiveawayScamHunter, the researchers were able to extract website and wallet addresses associated with the scams, ultimately uncovering 327 scam giveaway internet domains and 121 new scam-related cryptocurrency wallet addresses.
They also provided valuable insights into the workings of these scams, including how scammers target victims and the estimated number of victims scammed during the one-year study period.
The report estimated that over 365 victims were attacked by crypto scams during the period, leading to the loss of more than $872,000.
Furthermore, the researchers discovered that Twitter Lists, a networking tool available on the platform, have become a convenient vector for scammers to exploit due to their permissionless nature.
In order to identify which lists were involved in giveaway scams, the team trained a natural language processing tool on data from previously identified scams.
This enabled them to identify almost 100,000 instances of giveaway scam lists and collect data on previously unreported scam websites and wallets.
The researchers said they have shared their findings, along with the associated accounts, domains, and wallet addresses, with both Twitter and the crypto community.
However, according to their paper, 43.9% of the associated accounts remain active as of the paper’s publication on August 10.
“Our analysis also shows that 43.9% of spam accounts still remain active as of this writing, implying the urgent needs of detecting the spam accounts and preventing the dissemination on Twitter,” the report said.
Scammers Use AI to Develop New Scam Methods
The use of AI in detecting cryptocurrency giveaway scams on social media platforms comes as scammers and bad actors have already to develop new methods of fraud and deception.
By leveraging AI-powered tools, scammers can amplify their reach and create a seemingly loyal fanbase of thousands of people.
These fake accounts and interactions can be used to give the illusion of credibility and popularity to their scam projects.
Scammers may even use AI-driven chatbots or virtual assistants to engage with individuals, provide investment advice, promote fake tokens and initial coin offerings, or offer high-yield investment opportunities.
The use of AI can also challenge social proof-of-work, which assumes that crypto projects with greater and more loyal followings online must be legitimate.
One example of how scammers are using AI is through the use of “pig butchering” scams. AI instances can spend several days befriending someone, usually an elderly or vulnerable person, only to end up scamming them.